Dataset Information

Accuracy of random-forest-based imputation of missing data in the presence of non-normality, non-linearity, and interaction.

ABSTRACT:

Background

Missing data are common in statistical analyses, and imputation methods based on random forests (RF) are becoming popular for handling missing data especially in biomedical research. Unlike standard imputation approaches, RF-based imputation methods do not assume normality or require specification of parametric models. However, it is still inconclusive how they perform for non-normally distributed data or when there are non-linear relationships or interactions.

Methods

To examine the effects of these three factors, a variety of datasets were simulated with outcome-dependent missing at random (MAR) covariates, and the performances of the RF-based imputation methods missForest and CALIBERrfimpute were evaluated in comparison with predictive mean matching (PMM).

Results

Both missForest and CALIBERrfimpute have high predictive accuracy but missForest can produce severely biased regression coefficient estimates and downward biased confidence interval coverages, especially for highly skewed variables in nonlinear models. CALIBERrfimpute typically outperforms missForest when estimating regression coefficients, although its biases are still substantial and can be worse than PMM for logistic regression relationships with interaction.

Conclusions

RF-based imputation, in particular missForest, should not be indiscriminately recommended as a panacea for imputing missing data, especially when data are highly skewed and/or outcome-dependent MAR. A correct analysis requires a careful critique of the missing data mechanism and the inter-relationships between the variables in the data.

SUBMITTER: Hong S

PROVIDER: S-EPMC7382855 | biostudies-literature | 2020 Jul

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Accuracy of random-forest-based imputation of missing data in the presence of non-normality, non-linearity, and interaction.

Hong Shangzhi S Lynn Henry S HS

BMC medical research methodology 20200725 1

<h4>Background</h4>Missing data are common in statistical analyses, and imputation methods based on random forests (RF) are becoming popular for handling missing data especially in biomedical research. Unlike standard imputation approaches, RF-based imputation methods do not assume normality or require specification of parametric models. However, it is still inconclusive how they perform for non-normally distributed data or when there are non-linear relationships or interactions.<h4>Methods</h4> ...[more]

PMID: 32711455

Similar Datasets

Project description:BackgroundIn the pursuit of a better understanding of biodiversity, evolutionary biologists rely on the study of phylogenetic relationships to illustrate the course of evolution. The relationships among natural organisms, depicted in the shape of phylogenetic trees, not only help to understand evolutionary history but also have a wide range of additional applications in science. One of the most challenging problems that arise when building phylogenetic trees is the presence of missing biological data. More specifically, the possibility of inferring wrong phylogenetic trees increases proportionally to the amount of missing values in the input data. Although there are methods proposed to deal with this issue, their applicability and accuracy is often restricted by different constraints.ResultsWe propose a framework, called PhyloMissForest, to impute missing entries in phylogenetic distance matrices and infer accurate evolutionary relationships. PhyloMissForest is built upon a random forest structure that infers the missing entries of the input data, based on the known parts of it. PhyloMissForest contributes with a robust and configurable framework that incorporates multiple search strategies and machine learning, complemented by phylogenetic techniques, to provide a more accurate inference of lost phylogenetic distances. We evaluate our framework by examining three real-world datasets, two DNA-based sequence alignments and one containing amino acid data, and two additional instances with simulated DNA data. Moreover, we follow a design of experiments methodology to define the hyperparameter values of our algorithm, which is a concise method, preferable in comparison to the well-known exhaustive parameters search. By varying the percentages of missing data from 5% to 60%, we generally outperform the state-of-the-art alternative imputation techniques in the tests conducted on real DNA data. In addition, significant improvements in execution time are observed for the amino acid instance. The results observed on simulated data also denote the attainment of improved imputations when dealing with large percentages of missing data.ConclusionsBy merging multiple search strategies, machine learning, and phylogenetic techniques, PhyloMissForest provides a highly customizable and robust framework for phylogenetic missing data imputation, with significant topological accuracy and effective speedups over the state of the art.

Project description:BackgroundNon-additive interactions among genes are frequently associated with a number of phenotypes, including known complex diseases such as Alzheimer's, diabetes, and cardiovascular disease. Detecting interactions requires careful selection of analytical methods, and some machine learning algorithms are unable or underpowered to detect or model feature interactions that exhibit non-additivity. The Random Forest method is often employed in these efforts due to its ability to detect and model non-additive interactions. In addition, Random Forest has the built-in ability to estimate feature importance scores, a characteristic that allows the model to be interpreted with the order and effect size of the feature association with the outcome. This characteristic is very important for epidemiological and clinical studies where results of predictive modeling could be used to define the future direction of the research efforts. An alternative way to interpret the model is with a permutation feature importance metric which employs a permutation approach to calculate a feature contribution coefficient in units of the decrease in the model's performance and with the Shapely additive explanations which employ cooperative game theory approach. Currently, it is unclear which Random Forest feature importance metric provides a superior estimation of the true informative contribution of features in genetic association analysis.ResultsTo address this issue, and to improve interpretability of Random Forest predictions, we compared different methods for feature importance estimation in real and simulated datasets with non-additive interactions. As a result, we detected a discrepancy between the metrics for the real-world datasets and further established that the permutation feature importance metric provides more precise feature importance rank estimation for the simulated datasets with non-additive interactions.ConclusionsBy analyzing both real and simulated data, we established that the permutation feature importance metric provides more precise feature importance rank estimation in the presence of non-additive interactions.

Project description:BackgroundEpistatic miniarray profiling (E-MAPs) is a high-throughput approach capable of quantifying aggravating or alleviating genetic interactions between gene pairs. The datasets resulting from E-MAP experiments typically take the form of a symmetric pairwise matrix of interaction scores. These datasets have a significant number of missing values - up to 35% - that can reduce the effectiveness of some data analysis techniques and prevent the use of others. An effective method for imputing interactions would therefore increase the types of possible analysis, as well as increase the potential to identify novel functional interactions between gene pairs. Several methods have been developed to handle missing values in microarray data, but it is unclear how applicable these methods are to E-MAP data because of their pairwise nature and the significantly larger number of missing values. Here we evaluate four alternative imputation strategies, three local (Nearest neighbor-based) and one global (PCA-based), that have been modified to work with symmetric pairwise data.ResultsWe identify different categories for the missing data based on their underlying cause, and show that values from the largest category can be imputed effectively. We compare local and global imputation approaches across a variety of distinct E-MAP datasets, showing that both are competitive and preferable to filling in with zeros. In addition we show that these methods are effective in an E-MAP from a different species, suggesting that pairwise imputation techniques will be increasingly useful as analogous epistasis mapping techniques are developed in different species. We show that strongly alleviating interactions are significantly more difficult to predict than strongly aggravating interactions. Finally we show that imputed interactions, generated using nearest neighbor methods, are enriched for annotations in the same manner as measured interactions. Therefore our method potentially expands the number of mapped epistatic interactions. In addition we make implementations of our algorithms available for use by other researchers.ConclusionsWe address the problem of missing value imputation for E-MAPs, and suggest the use of symmetric nearest neighbor based approaches as they offer consistently accurate imputations across multiple datasets in a tractable manner.

Dataset Information

Accuracy of random-forest-based imputation of missing data in the presence of non-normality, non-linearity, and interaction.

Background

Methods

Results

Conclusions

Publications

Accuracy of random-forest-based imputation of missing data in the presence of non-normality, non-linearity, and interaction.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets